Localization and Surface Characterization byZhurong Mars Rover at Utopia PlanitiaLiang Ding ( [email protected] )
State Key Laboratory of Robotics and System, Harbin Institute of Technology https://orcid.org/0000-0002-8351-5178Ruyi Zhou
State Key Laboratory of Robotics and System, Harbin Institute of Technology https://orcid.org/0000-0002-1854-9451Tianyi Yu
Beijing Aerospace Control CenterHaibo Gao
Harbin Institute of TechnologyHuaiguang Yang
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyJian Li
Beijing Aerospace Control CenterYe Yuan
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyChuankai Liu
Beijing Aerospace Control CenterJia Wang
Beijing Aerospace Control CenterYuyan Zhao
Institute of Geochemistry Chinese Academy of Sciences https://orcid.org/0000-0003-2551-5425Zhengyin Wang
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyXiyu Wang
Center for Lunar and Planetary Sciences, Institute of Geochemistry, Chinese Academy of Science;University of Chinese Academy of SciencesGang Bao
Center for Lunar and Planetary Sciences, Institute of Geochemistry, Chinese Academy of Science;University of Chinese Academy of SciencesZongquan Deng
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyLan Huang
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyNan Li
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyXiaofeng Cui
Beijing Aerospace Control CenterXiming He
Beijing Aerospace Control CenterYang Jia
Beijing Institute of Spacecraft System EngineeringBaofeng Yuan
Beijing Institute of Spacecraft System EngineeringGuangjun Liu
Department of Aerospace Engineering, Ryerson UniversityHui Zhang
Beijing Aerospace Control CenterRui Zhao
Beijing Aerospace Control CenterZuoyu Zhang
Beijing Aerospace Control CenterZiqing Cheng
Beijing Aerospace Control CenterFan Wu
Beijing Aerospace Control CenterQian Xu
Beijing Aerospace Control CenterHao Lu
Beijing Aerospace Control CenterLutz Richter
Large Space Structures (LSS)Zhen Liu
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyFuliang Niu
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyHuanan Qi
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyShu Li
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyWenhao Feng
State Key Laboratory of Robotics and System, Harbin Institute of Technology
Chaojie Yang State Key Laboratory of Robotics and System, Harbin Institute of Technology
Baichao Chen Beijing Institute of Spacecraft System Engineering
Zhaolong Dang Beijing Institute of Spacecraft System Engineering
Mingming Zhang Center for Lunar and Planetary Sciences, Institute of Geochemistry, Chinese Academy of Science;
University of Chinese Academy of SciencesLichun Li
Beijing Aerospace Control CenterXiaoxue Wang
Beijing Aerospace Control CenterZhao Huang
Beijing Aerospace Control CenterJitao Zhang
Beijing Aerospace Control CenterHongjun Xing
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyGuanyu Wang
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyLizhou Niu
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyPeng Xu
State Key Laboratory of Robotics and System, Harbin Institute of TechnologyWenhui Wan
State Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, ChineseAcademy of SciencesKaichang Di
State Key Laboratory of Remote Sensing Science, Institute of Remote Sensing and Digital Earth, ChineseAcademy of Sciences, Beijing 100101, China
Article
Keywords: Mars rover, Mars, Utopia Planitia, Zhurong rover
Posted Date: September 24th, 2021
DOI: https://doi.org/10.21203/rs.3.rs-836162/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License
1
Localization and Surface Characterization by Zhurong Mars Rover at Utopia Planitia
L. Ding1*†, R. Zhou1†, T. Yu2†, H. Gao1*, H. Yang1*, J. Li2, Y. Yuan1, C. Liu2, J. Wang2, Y. Zhao3,4, Z.
Wang1, X. Wang3,5, G. Bao3,5, Z. Deng1, L. Huang1, N. Li1, X. Cui2, X. He2, Y. Jia6, B. Yuan6, G. Liu7, H.
Zhang2, R. Zhao2, Z. Zhang2, Z. Cheng2, F. Wu2, Q. Xu2, H. Lu2, L. Richter8, Z. Liu1, F. Niu1, H. Qi1, S.
Li1, W. Feng1, C. Yang1, B. Chen6, Z. Dang6, M. Zhang3,5, L. Li2, X. Wang2, Z. Huang2, J. Zhang2, H. 5
Xing1, G. Wang1, L. Niu1, P. Xu1, W. Wan9, K. Di9
1State Key Laboratory of Robotics and System, Harbin Institute of Technology; Harbin 150080, China.
2Beijing Aerospace Control Center; Beijing 100094, China.
3Center for Lunar and Planetary Sciences, Institute of Geochemistry, Chinese Academy of Science,
Guiyang 550081, China. 10
4CAS Center for Excellence in Comparative Planetology, Hefei 230026, China.
5University of Chinese Academy of Sciences, Beijing 100049, China.
6Beijing Institute of Spacecraft System Engineering; Beijing 100094, China.
7Department of Aerospace Engineering, Ryerson University; Toronto, ON M5B 2K3, Canada.
8Large Space Structures (LSS), Hauptstrasse 1e, Eching, Germany. 15
9State Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese
Academy of Sciences; Beijing 100101, China.
*Corresponding author. Email: [email protected] (L. D.), [email protected] (H. G.),
[email protected] (H. Y.).
†These authors contributed equally to this work 20
2
China’s first Mars rover, Zhurong, has successfully touched down on the southern Utopia Planitia of
Mars at 109.925° E, 25.066° N, and since performed cooperative multiscale investigations with the
Tianwen-1 orbiter. Here we present primary localization and surface characterization results based
on complementary data of the first 60 sols. The Zhurong rover has traversed 450.9 m southwards
over a flat surface with mild wheel slippage (less than 0.2 in slip ratio). The encountered crescent-5
shaped sand dune indicates a NE-SW local wind direction, consistent with larger-range remote-
sensing observations. Soil parameter analysis based on terramechanics indicates that the topsoil has
high bearing strength and cohesion, and its equivalent stiffness and internal friction angle are ~1390-
5872 kPa∙m-n and ~21°-34° respectively. Rocks observed strewn with dense pits, or showing layered
and flaky structures, are presumed to be involved in physical weathering like severe wind erosion 10
and potential chemical weathering processes. These preliminary observations suggest great potential
of in-situ investigations by the scientific payload suite of the Zhurong rover in obtaining new clues of
the region’s aeolian and aqueous history. Cooperative investigations using the related payloads on
both the rover and the obiter could peek into the habitability evolution of the northern lowlands on
Mars. 15
The Mars Exploration Probe Tianwen-1 lander (including the Zhurong rover) has landed on the Utopia
Planitia of Mars on May 15, 20211 and successfully fulfilled the goals of ‘orbiting, landing, and roving’ on
Mars at China’s first independent attempt. The Zhurong rover (1230×830×540 mm, 240 kg; Extended Data
Fig. 1), a six-wheeled solar-powered robot with active suspension, working in high locomotive
performance (Extended Data Table 1) and carrying six scientific payloads2-7, becomes one of the most 20
powerful rovers to date that ever landed on the Northern Plain of Mars. Zhurong has been roving and
conducting in-situ investigations for more than the designed 90-sols primary mission, collaborating with
the orbiter to collect complementary scientific data at different perception scales and precisions for five
primary scientific objectives7. Scientific data of ~7 Gbit has been relayed back via the Tianwen-1 orbiter
3
and is currently under calibration. Here we present preliminary results in terms of the localization and
surface characterization of the landing site based on the first 60-sols cooperative observations from the
Zhurong rover and the Tianwen-1 orbiter.
The landing site location was refined to 109.925° E, 25.066° N using the descent image sequence and
mapped to the orbiter-derived digital elevation map (DEM) at an elevation of –4099.8 m. It was further 5
confirmed by the High-Resolution Imaging Science Experiment (HiRISE) onboard the Mars
Reconnaissance Orbiter (MRO; Extended Data Fig. 2), in which the lander and rover can be identified. The
landing site (Fig. 1a) lies in the Late Hesperian lowland unit, dated to ~3.32–3.36 Ga8. The region contains
hundreds of superposed pedestal-crater forms, a thumbprint terrain, topographically-subdued wrinkle ridges,
and narrow grabens northeast of Alba Mons9. We highlight the local area in a circular shape that might be 10
within reach of the rover (centred on 109.925° E, 24.980° N; diameter = 20 km; Fig. 1b). The elevation of
the local area gradually increasing due south from -4219 to -3989 m, with an average elevation of -4095 m
and a standard deviation of 42.11 m (based on the Mars Orbiter Laser Altimeter (MOLA) DEM 128PPD;
463 m·pixel-1). About 99% of the area consist of slope less than 2.3°, and relatively large slopes are
primarily distributed at the rim of impact craters (Fig. 1c). Therefore, the area is flat and may facilitate the 15
long-distance rovering and exploration of the Zhurong rover.
Zhurong has traveled 450.9 m due south during the first 60 sols (Fig. 2a), with waypoints calculated
by cross-site visual localization10, and conducted several in-situ investigations (Extended Data Table 2). An
off-line dynamic locomotion simulation11 (Extended Data Fig. 3) implemented on the Rover Simulation
based on Terramechanics Dynamics (RoSTDyn)12 is used to evaluate and verify the safety and efficiency of 20
the planned path. Compared with the visual localization postures, the simulations of planned traverses
deviated by ≤ 2%, and the average azimuth error was ≤ 1°. Due to the inevitable wheel slippage, the
planned destination was not always accurately reached, particularly when traversing through drifting soils.
The wheel slip ratio (denoted by s) over each continuous steady traverse is calculated using the Guidance,
4
Navigation and Control System (GNC)-derived rover velocity and the wheel angular velocity derived from
the encoder (see Methods). The distribution of the wheel slip ratio along the route (Fig. 2b) shows that the
Zhurong rover primarily operates under mild wheel slippage (s < 0.2). The average slip ratio of a steady
climb of ~0.55° was 0.056, consistent with the elevation increase of 4.34 m during the 60-sol journey (Fig.
2c). The largest average slip ratio of 0.22 occurred on sol 55 while Zhurong climbed a slope of ~1.14°. 5
Occasionally, the rover suffers wheel skids (s < 0) on localized down slopes, e.g., traversing through a
small down-slope area of ~-3° on sol 34 before approaching a crescent-shaped dune. Wheel tracks in
periodic textures (Extended Data Fig. 4a) left by the Zhurong rover also provide clues on the severity of
wheel slippage in a more detailed view. By extracting the wheel track unit (see Methods), the average
wheel slip ratio on legible parts of the wheel track (being created after Zhurong traversed from waypoints X 10
to A; Extended Data Fig. 4b) was calculated to be 0.05, consistent with telemetry-derived results.
The first 360° panorama is stitched from twelve images taken by the Navigation and Terrain Camera
(NaTeCam) (Fig. 3a). The panorama shows an overview of gentle topography at the landing site, with
major surface geological features including aeolian dunes, small craters, and rocks. One distinct feature is
that the distribution of several aeolian dunes with ripple-like structures, which can be reached and 15
investigated by the rover. Most dunes in this area are bright-toned and crescent-shaped, with windward
slopes (the protruding side) oriented approximately north-eastwards, indicating a NE-SW local wind
direction (Extended Data Fig. 2b). The first dune encountered by the Zhrong rover on sol 50 is 40 m in
length, 8 m in width, and 0.6 m in height (Fig. 3b), with sands in two different colors covered the dune
surface. Several seagull-shaped deposits are usually formed from two merged dunes, and one approximate 20
to a line as shown in the enlarged view (Extended Data Fig. 2b). Further scientific observations from the
orbit are needed to determine whether these dunes are currently active. A fresh mini crater (~0.95 m in
diameter) beneath the lander created by the landing plume was observed in the NaTeCam image, providing
clues on a shallow subsurface layer (Fig. 3c). Gravels (1-4 cm in diameter), splashed out around the lander
5
struts or settled on the crater rim, exhibit a dark-brown tone and in sharp contrast to the semi-buried ones
by the dust.
The mechanical properties of the surface soils (Fig. 3d) are estimated using the lugged wheels of
Zhurong as test devices. Using stereo images taken by the rear Hazard-avoidance Camera (HazCam), the
soil surface deformation by lugged wheels can be reconstructed and the wheel sinkage can be estimated by 5
extracting the wheel track profile, as it was done for the Chang’E-3 mission13. In legible parts of the wheel
tracks (Extended Data Fig. 4d), the equivalent wheel sinkage (the sum of the contribution of the lug and the
drum-shaped wheel) was approximated to be 10 mm. However, most wheel tracks are incomplete without
well-trimmed contours (Extended Data Fig. 4e). The wheel tracks are formed in a way that the 5-mm-high
wheel lugs almost completely immerse into the soil, with the wheel rim several millimetres above the 10
surface (as observed in the images on sol 12 taken by a wireless camera deployed on the ground; Extended
Data Fig. 4c). Therefore, the equivalent wheel sinkage was estimated to be about 5 mm. Wheel sinkage is
sometimes interrupted by protruding gravels on the surface , leading to smaller values (about 2 mm) or
going larger to approximately 15 mm when the wheel rim sinks below the surface (Extended Data Fig. 4f).
Compared with Yutu or Yutu-2 rovers of the lunar exploration missions, the load on each wheel of 15
Zhurong (~148.8 N) is much larger than that on the wheels of lunar rovers (~36.5 N). However, the
corresponding wheel sinkage during the Zhurong traverses was not more significant, indicating a greater
bearing strength of the Martian soil than that of the lunar regolith. According to the terramechanics model14
for the lugged wheels, characteristic curves of the sinkage exponent (denoted by n) and the equivalent
stiffness (denoted by Ks) were predicted under different sinkage conditions (Fig. 4a). We set the upper 20
bound of n to 1.0 (i.e., the typical value of lunar regolith15). As n increases, the corresponding soil soften
and results in a sharp increase in wheel sinkage. Also, the lower bound of n is set to 0.7 to maintain Ks
within the reasonable range because the Ks decreases with a decresing n value. Considering soil
characteristics within curves corresponding to a wheel sinkage of 2-5 mm and representative terrestrial soil
6
types, the equivalent stiffness Ks was estimated to be 1390-5872 kPa∙m-n, shown as the rectangular region
in Fig. 4a.
Using the wheel-terrain interaction model, the characteristic shearing curves under different driving
torques (4.0-8.0 N∙m) were predicted (Fig. 4b). Transformed from the wheel driving motor current in
moving states, the average wheel driving torque during each continuous traverse during sols 23-34 ranged 5
3.7-7.0 N∙m (Extended Data Table 3). The maximum wheel driving torque reached 8-10 N∙m during some
sampling periods, although the corresponding slip ratio increased accordingly. Therefore, 8 N∙m was taken
as the upper bound of the driving torque. Taking the cohesion (denoted by c) of 6 kPa (the upper bound;
close to the maximum soil cohesion from North Gower clayey16 loam in a reasonable range), and 1.5 kPa
(the lower bound; the minimum cohesion of representative soil types on Earth) respectively, the internal 10
friction angle (denoted by φ) was limited to ~21°-34° in the area enclosed by these curves. The friction
characteristics of the soil at the Tianwen-1 site are inferior to those of the InSight, Phoenix, and Mars
Pathfinder landing sites. Meanwhile, the soil cohesion of the Tianwen-1 site is relatively large, resulting in
soil adhering to the wheel surface during the traverse (Extended Data Fig. 4c). Compared with the soil
shearing properties in other Mars missions17-23, the in-situ results of Tianwen-1 are within the envelop 15
region of others and closest to that of the Viking 1, and Curiosity sites (Fig. 4c).
The flat nature of the landing area is occasionally interrupted by small craters. More than 2000 craters
were identified on the HiRISE image (spatial resolution ~0.3 m) within the circular area surrounding the
Tianwen-1 site (centred on 109.925° E, 25.048° N; diameter = 3 km; Extended Data Fig. 5a), and a size-
frequency distribution analysis (Extended Data Fig. 5b) was conducted. The crater diameter in the area 20
ranges from 1.1 and 287.6 m, with a mean diameter of 10.6 m. About 79% of these craters are ≤ 10 m in
diameter, and most are likely secondary craters. Relatively large craters (≥ 200 m in diameter) (as C1, C2,
C3 in Extended Data Fig. 5a) show clear impact structures but had broken or degraded rims. Craters that
7
Zhurong has encountered are relatively small (< 10 m in diameter) (Extended Data Fig. 5c, d) in a
depression-shape, surrounded by dark rocks (probably ejecta). Aeolian deposits within the craters may
indicate that the craters may be subject to long-term weathering.
The surface is primarily free of large boulders but is scattered with small rocks and clasts bearing
distinct features (Extended Data Fig. 6). Most rocks are fine-grained in texture and angular to subangular, 5
with low roundness in morphology. Some rocks with pitted surfaces show similar morphology to igneous
rocks as observed in previous missions (e.g., by the Viking-1 lander and the Spirit rover)24,25, which are
proposed to form via brine-related dissolution processes under cold environments25,26. Similar rock features
may provide insights into the climate and geological processes of the Tianwen-1 site but require detailed
investigations by the scientific payloads of the rover. Some rock chucks also show flaky texture, similar to 10
the rock targets observed at Gusev Plain, such as “Mimi”27. The flake textures might be related to aqueous
alterations that dew water goes through insolation cracking to flake the rocks, and brine and salt may work
to cement these flakes28. In addition, rocks with grooves and etchings on the windward sides are ubiquitous
on the landing site, usually interpreted as ventifacts resulting from intense wind erosion with aeolian
particles29-31. Therefore, the rock textures observed at the site so far may indicate both presences of 15
physical weathering (e.g., impact sputtering, wind erosion, and potential freeze-thaw weathering) and
aqueous interactions, involving salt and brine activities. These rock and soil targets provide excellent
opportunities to peek into the aqueous history and climate evolution of the northern lowlands and shed light
on the habitability evolution of Mars.
References 20
1. Mallapaty S. Nature 593, 323-324 (2021).
2. Li, C. et al. Scientific objectives and payload configuration of China’s first mars exploration mission. J.
Deep Space Explor. 5, 5 (2018).
8
3. Liang, X. et al. The navigation and terrain cameras on the Tianwen-1 Mars rover, Space Sci. Rev. 217,
37 (2021).
4. Zhou, B. et al. The Mars rover subsurface penetrating radar onboard China’s Mars 2020 mission, Earth
Planet. Phys. 4, 345-354 (2020).
5. Peng, Y. et al. Overview of the Mars climate station for Tianwen-1 mission, Earth Planet. Phys. 4, 5
371-383 (2020).
6. Du, A. et al. The Chinese Mars rover fluxgate magnetometers, Space Sci. Rev. 216, 135 (2020).
7. Li, C. et al. Design and realization of Chinese Tianwen-1 energetic particle analyzer, Space Sci. Rev.
217, 26 (2020).
8. Wu, B. et al. Characterization of the candidate landing region for Tianwen-1-China’s first mission to 10
Mars, Earth Space Sci. 8, e2021EA001670 (2021).
9. Tanaka, K.L. et al. Geologic map of Mars: U.S. Geological Survey Scientific Investigations Map 3292,
(US Geological Survey, 2014).
10. Liu, S. et al. High precision landing site mapping and rover localization for Chang’e-3 mission. Sci.
China Phys. Mech. Astron. 58, 1-11 (2015). 15
11. Yang, H. et al. paper presented in the 2018 IEEE International Conference on Robotics and
Automation, Brisbane, QLD, Australia, 21 to 25 May 2018.
12. Li, W., Ding, L., Gao, H., Deng, Z., Li, N. ROSTDyn: Rover simulation based on terramechanics and
dynamics. J. Terramech. 50, 199-210 (2013).
13. Gao, Y., Spiteri, C., Li, C., Zheng, Y. Lunar soil strength estimation based on Chang’E-3 images. Adv. 20
Space Res. 58, 1893-1899 (2016).
14. Ding, L. et al. Interaction mechanics model for rigid driving wheels of planetary rovers moving on
sandy terrain with consideration of multiple physical effects. J. Field Robot. 32, 827-859 (2015).
9
15. French, B. M., Heiken, G., Vaniman D. Lunar Sourcebook: A User’s Guide to the Moon (Cambridge
University Press, Cambridge, 1991)
16. Wong, J.Y. Theory of Ground Vehicles (John Wiley & Sons, New Jersey, 2008).
17. Henry, J., Bruce, M. Viking landing sites, remote-sensing observations, and physical properties of
Martian surface materials. Icarus 81, 164-184 (1989). 5
18. Rover team, Characterization of the martian surface deposits by the Mars Pathfinder rover, Sojourner.
Science 278, 1765-1768 (1997).
19. Arvidson, R. E. et al. Localization and physical properties experiments conducted by Spirit at Gusev
Crater. Science 305, 821-824 (2004).
20. Arvidson, R. E. et al. Localization and physical property experiments conducted by Opportunity at 10
Meridiani Planum. Science 306, 1730-1733 (2004).
21. Arvidson, R. E. et al. Terrain physical properties derived from orbital data and the first 360 sols of Mars
Science Laboratory Curiosity rover observations in Gale Crater. J. Geophys. Res. Planets. 119, 1322-
1344 (2013).
22. Any, S. et al. Phoenix soil physical properties investigation. J. Geophys. Res. 114, E00E0 (2009). 15
23. Golombek, M. et al. Geology of the InSight landing site on Mars. Nat. Commun. 11, 1014 (2020).
24. Mutch, T. A. et al. The surface of Mars: The View from the Viking 1 lander. Science 193, 4255, 791-
801 (1976)
25. Head, J. M., Kreslavsky, A., Marchant, D. R. Pitted rock surfaces on Mars: a mechanism of formation
by transient melting of snow and ice. J. Geophys. Res. 116, E09007 (2011). 20
26. Cabrol, N. A. et al. Aqueous processes at Gusev crater inferred from physical properties of rocks and
soils along the Spirit traverse. J. Geophys. Res. Planets. 111, E02S20 (2006).
10
27. Arvidson, R. E., et al. Overview of the Spirit Mars Exploration Rover Mission to Gusev Crater:
Landing site to Backstay Rock in the Columbia Hills, J. Geophys. Res., 111, E02S01 (2006).
28. Thomas, M., Clarke, J., Pain, C. Weathering, erosion and landscape processes on Mars identified from
recent rover imagery, and possible Earth analogues. Aust. J. Earth Sci. 52, 365-378 (2005).
29. Greeley, R. et al. Gusev Crater: Wind-related features and processes observed by the Mars Exploration 5
Rover Spirit. J. Geophys. Res.-Atmos. 111, E02S09 (2006).
30. Thomson, B. J., Bridges, N. T., Greeley, R. Rock abrasion features in the Columbia Hills, Mars. J.
Geophys. Res. Planets. 113, E08010 (2008).
31. Bridges, N. T. et al. Ventifacts at the Pathfinder landing site. J. Geophys. Res. Planets. 104, 8595-8615
(1999). 10
11
Fig. 1. Regional topographic map of the Tianwen-1 landing site. a, The Tianwen-1 landing site is at the
southern Utopia Planitia, Mars. The map shows the major physiographic features around the Tianwen-1
touch-down spot and the landing sites of previous missions, including the Viking Lander 2 (VL2), the
InSight Lander, Mars Science Laboratory (MSL), the Mars Exploration Rover (MER-A) Spirit rover. The 5
base map is a portion of the MOLA shaded-relief topographic map of Mars with elevations for the geoid. b,
MOLA-derived elevation map of the Tianwen-1 surrounding region in a circular shape. The 20 km
diameter circular map is centered at 109.925° E, 24.980° N, about 10 km south of the Tianwen-1 landing
site. The average elevation is -4095 m in the region. c, The slope map of the Tianwen-1 surrounding region
is based on the MOLA data (baseline 926 m). About 99% of the topography is flat (slopes < 2.3°), and the 10
area with more significant slopes (> 2.3°) primarily distributes on the rim of the impact craters.
12
Fig. 2. The routing path of the Zhurong rover and the associated wheel slippage for the first 60 sols.
a, The routing path of the Zhurong rover. The green dots represent starting waypoint on each sol. The rover
has passed through 55 waypoints during the first 60-sols operation. The base image is a high-resolution
(~0.7 m/pixel) digital orthophoto map (DOM) generated from the High Resolution Imaging Camera 5
(HiRIC) images taken by the Tianwen-1 orbiter. b, The box plot of wheel slip ratios of the Zhurong rover.
Only the sols when Zhurong moves in high-efficiency mode (from sol 23 on) are plotted. Each box
represents the distribution of wheel slip ratios over continuous traverses experienced in that sol. Traveling
distances on sols 42-48 are short, causing relatively large locomotion measurement errors and resulting in
the slip ratio in disagreement with the elevation trend. c, The elevation profile along the traversed path 10
during the first 60 sols. The blue dots represent the rover elevation of starting waypoints on each sol, while
the yellow dots represent the elevation within the waypoints. The rover elevation of waypoint X (the initial
waypoint on the surface) is taken as the baseline (elevation of 0), and the elevation varies from 0 (on sol
10) to 4.34 (on sol 60) m.
15
13
Fig. 3. The geological features at the landing site. a, The first 360° panorama of the Tianwen-1 landing
site taken by the NaTeCam. The image is stitched by 12 images taken at 30° intervals by the NaTeCam on
sol 6. The Zhurong rover is heading southwards, and the parachute and backshell are located a few hundred
meters to the southwest of the lander. Two jet-wasted traces reveal darker materials below the topsoil. On 5
the southeast of the lander, there are several bright sand dunes, a small crater surrounded by several dark-
colored rocks, probably ejected from the crater. b, The first aeolian dune that Zhurong rover has
encountered. The windward slope is relatively gentler than the leeward one. The image was taken on sol
50. c, A mini crater (~0.95 m in diameter) right underneath the lander formed by the thrust engine plume of
the Tianwen-1 probe during landing. The crater reveals two distinct rock fragments. The one with a dark-10
brown tone contrasts sharply with the bright-tone clasts semi-buried in the soils and dust. d, Soil surface
scattered with bright-toned clasts (green arrows) and small-sized dark-toned rocks (yellow arrows).
14
Fig. 4. Analysis of soil mechanical parameters at the Zhurong landing site. a, Soil bearing
characteristic curves under different wheel sinkage for Zhurong’s wheels and the bearing parameters of
typical soil samples on Earth. The soil bearing properties at the Tianwen-1 landing site are close to
representative soil samples on Earth, such as the LETE sand (Wong) (red triangle), Grenville loam (blue 5
square), and North Gower clayey loam (Wong) (green pentagon). The bearing parameters of soil samples at
the Tianwen-1 site are presumed to be within the area (in light orange) containing five soil samples it it.
The bearing parameters of typical soil samples on Earth are listed in Supplementary Table 1. b, Soil
shearing characteristic curves derived from rover driving torque for Zhurong’s wheels and the shearing
parameters of typical soil samples on Earth. The more significant driving torque used for traverse 10
represents stronger shearing resistance of the soil. The shearing parameters of typical soil samples on Earth
are listed in Supplementary Table 1. c, Soil shearing parameters of Tianwen-1 landing site compared with
that of other Mars landing sites. The shearing parameters of soil samples on other Mars landing sites are
listed in Supplementary Table 2.
15
15
Methods
Instruments and data description
The Zhurong rover is a six-wheeled solar-powered robot with active suspension. Benefited from its
active suspension structure32, the rover is not only able to work in basic wheeled movement modes but also
capable of novel gaits, such as a crab gait for cross-walking, a creeping gait for better slope climbing 5
capability, a wheel uplift gait to escape from wheels being stuck33, and a body uplift/settlement gait. The
strong terrain adaptability and resilient fault recovery capability allow the rover to access dangerous but
scientifically beneficial regions, such as sand dunes and crater rims, leading to significant scientific
discoveries. The lugged wheels on the Zhurong rover are used as devices for soil mechanical parameter
analysis based on terramechanics. Considering the gravity on Mars, the vertical load on each wheel is 10
estimated to be 148.8 N under a quasi-static state, based on the rover’s configuration and mass distribution.
The parameters of the drum-shaped wheel are in the Extended Data Table 1.
The data used in this study includes images from a NaTeCam, two HazCams, and a wireless camera,
alongside locomotion data from the onboard inertial measurement unit and wheel encoders.
The NaTeCam, one of the scientific payloads, is mounted on the mast of the Zhurong rover. It consists 15
of two optical systems with identical functions, performances, and interfaces. The parameters of the
NaTeCam are in the Supplementary Table 3. The NaTeCam is used for three-dimensional panoramic
imaging of the Mars surface, and to study topography and geological structure of the roving area. It takes
12 pairs of images in sequence to compose a 360° global perception.
The HazCam and wireless camera are engineering payloads onboard the Zhurong rover. There are two 20
pairs of separately installed HazCams at the front and back of the rover’s carriage. The parameters of the
HazCam are also in the Supplementary Table 3. The wireless camera is hidden at the bottom of the rover’s
carriage. It is equipped with WiFi components and has WiFi communication capabilities. The camera was
16
released at sol 17 to take a selfie of both the rover and the lander. There is also a WiFi receiving device on
the lander responsible for receiving WiFi image data and transmitting it back to Earth.
The locomotion data used in this study included the rover’s position (x, y and z) and posture (roll,
pitch and yaw) at the landing site local (LSL) coordinate frame, the rover’s velocities along three axes in its
coordinate frame, and the angular velocities of each of the wheels. The rover’s positions and postures are 5
derived from the inertial measurement unit onboard, and the wheels’ angular velocities are derived from the
angle recorded by wheel encoders. The LSL coordinate frame is a north-east-down right-handed coordinate
system with its origin at the first waypoint. Its z-axis points down in the local normal direction, the x-axis
points to the north pole, and the y-axis is orthogonal to the x- and z- axes. The rover coordinate frame is an
east-north-up right-handed local system, whose origin is at the rover’s centre, its x-axis points to the 10
forward direction of the rover, its z-axis points in the local normal direction, and its y-axis is orthogonal to
the x- and z-axes. These data were recorded onboard at a higher frequency, but the data transmitted to Earth
are only at a frequency of 1/15 Hz due to the communication channel restriction, with both recorded and
received timestamps.
Wheel slip ratio estimation based on telemetry data 15
The slip ratio s of a lugged wheel at each moment t is defined as follows:
( )( ) ( )( ) ( ) ( ) ( )( )( ) ( ) ( ) ( ) ( )( )
s s
s s
1 ,0 1
1 , 1 0
v t r t r t v t s ts t
r t v t r t v t s t
ω ω
ω ω
− ≥ ≤ ≤= − < − ≤ <
, (1)
where ω(t) is the angular velocity function, rs is the shearing radius function, and v(t) is the linear velocity
function. The wheel angular velocity is recorded by wheel encoders, and its linear velocity is derived based
on the rover’s linear velocity (recorded by the onboard inertia measurement unit) and the curvature of the 20
trajectory. The shearing radius sr can be computed as34:
s sr r hλ= + , (2)
17
where r is wheel radius, h is the lug height and sλ (
s1 0λ≤ ≤ ) is the lug coefficient determined by the
number of lugs and the internal friction angle of the soil. As Zhurong’s wheels are drum-shaped, its radius r
in equation (2) is set as a value between the largest radius and the smallest radius. Here, it is set as the mean
of wheel’s largest radius and its smallest radius. Besides, Zhurong’s wheels are with evenly arranged 5mm-
high lugs, thus the value of sλ is approximately 0.5 according to the experiment results in 34. 5
A slip ratio s value larger than zero indicates that the wheel has slipped; an s equal to 0 indicates that
the wheel has rolled without slipping or skidding; and an s less than 0 indicates that the wheel has skidded,
with |s| being the values of the skid ratio. Generally, a driving wheel slips when moving on flat terrain or
climbing up a slope, and may skid when moving down a slope33.
Slip ratio estimation based on wheel imprint 10
The slip ratio of a Zhurong rover wheel can be estimated from the imprint of the wheel. A clear and
intact wheel imprint with a longitudinal slip, as shown in Extended Data Fig. 4a, consists of a series of
trace units, whose width is denoted by px∆ , and each trace unit can be divided into two areas: the lug-
sheared area dug out by the lug itself and the hub-pressed area compressed by the wheel hub. Through
analysis of the formation mechanism of wheel track imprints with longitudinal slip, the relationship of the 15
average wheel slip ratio s and the width of the trace unit px∆ can be deduced as35:
( ) ( )s1 2 0 1ps x n r sπ= −∆ ≤ ≤ , (3)
where n is the number of lugs on the wheel. Therefore, s can be estimated by measuring px∆ .
The width of the trace unit px∆ is estimated using NaTeCam or rear HazCam images (Extended Data
Fig. 4b, c). An image processing method was introduced to extract px∆ with two key steps: view correction 20
and trace feature extraction. The original image is first corrected with the camera extrinsic matrix to a top-
view image. Then, each trace imprint unit is manually divided into independent areas, and trace imprint
18
information, such as the trace unit width, is acquired. The wheel slip ratio is then estimated from the trace
imprint using equation (3). When a series of continuous trace units has almost the same width, the average
trace unit widths are calculated to reduce the measurement error, and the average wheel slip ratio is
deduced to represent the wheel slippage experienced by Zhurong. Since only whole imprints of the rear
wheels are left behind the rover while the front and the middle wheel imprints are typically overlapped, the 5
slip ratio of the rear wheels was estimated to represent the rear wheel slippage at the respective moment.
Soil parameter analysis
Zhurong’s wheels were used as devices to analyze soil parameters based on wheel-soil interaction. For
a lugged rover wheel moving on the soil with an angular velocity ω, the wheel is applied with a vertical
load W and a resistance force fDP from the vehicle suspension, as well as a driving torque T at the wheel 10
rotational axis by an actuator. The terrain interacts with the wheel circumference in the contact region,
which corresponds to the angle divided into two parts: the entrance angle θ1 from the vertical at which the
wheel first makes contact with the soil, and the exit angle θ2 from the vertical, at which the wheel loses
contact with the soil. In the wheel–terrain interaction region (θ1+θ2), the continuous normal stress σ to
support the wheel, and the shearing stress τ due to the relative movement are exerted on the wheel surface, 15
as shown in Extended Data Fig. 7a. The point of maximum stress is denoted as θm, according to which the
stress region is divided into a forward part (σ1, τ1), corresponding to the angle from θ1 to θm, and a rear part
(σ2, τ2), corresponding to the angle from θm to θ2.
The soil bearing parameters are closely related to the wheel-soil interaction in the normal direction.
Considering the shape of the wheel surface, the distributed normal stress36 along the wheel circumference 20
deduced from the Reece-Wong model37 is
19
c1 s 1 m 1
c 22 s 1 1 m 1 2 m
m 2
( ) ( ) (cos cos ) ( )
( ) ( ) cos ( ) cos ( )
n n
n
n
kk r
b
kk r
b
σ θ θ θ θ θ θ
θ θσ θ θ θ θ θ θ θ θθ θ
ϕ
ϕ
= + − ≤ < − = + − − − ≤ < −
, (4)
where b is the width of the wheel, rs is the equivalent wheel radius calculated on equation (2), ck is the
cohesive modulus of the soil, kϕ is the frictional modulus of the soil, and n is a soil sinkage exponent that
can be represented by a linear function of slip ratio s as follows:
0 1n n n s= + , (5) 5
where n0 represents the static sinkage exponent and n1 represents the dynamic sinkage resulting from wheel
slippage. The values of n0 and n1 are determined based on experiments, and the introduction of the linear
soil sinkage exponent leads to a high affinity for fitting the slip-sinkage phenomenon of lugged wheels14.
The entrance angle θ1 is a function of wheel sinkage z as:
1 arccos[( ) / ]r z rθ = − , (6) 10
and the exit angle θ2 is computed as:
2 3 1cθ θ= , (7)
where c3 is a coefficient of the wheel-terrain interaction angle, generally assumed to be 038. The exit angle
θ2 is used to compute the stress caused by the rebound of the soil (stress integration from θm to θ2), which is
usually negligibly small and is taken as zero in the calculation. 15
The maximum stress angle θm is computed as:
m 1 2 1( )c c sθ θ= + , (8)
where c1 and c2 are coefficients of the wheel–terrain interaction angle. In the calculation, c1 and c2 are set to
0.5 and 0, respectively, because it is reasonable to assume that the angular location of maximum stress θm
occurs midway between θ1 and θ239. 20
20
The tangential stress τ(θ) is computed as
{ }' '
s 1 1( ) ( ( ) tan ) 1 exp( ( ) (1 )(sin sin ) / )c r s kτ θ σ θ ϕ θ θ θ θ = + × − − − − − − , (9)
where c is the cohesion of the soil, ϕ is the internal friction angle, k is the shearing deformation modulus, rs
is the equivalent shearing radius calculated on equation (2), and 1θ ′ is the equivalent entrance angle. With
the lug effect, the shearing stress occurs on the surface of the soil sticking to the wheel circumference due 5
to the wheel lugs instead of the wheel outer cylinder surface. The equivalent entrance angle 1θ ′ of a lugged
wheel14 is computed as
1 arccos[( ) / ( )]r z r hθ ′ = − + . (10)
When θ approaches mθ , the corresponding normal stress and tangential stress approach their maximum as
c
m m 1( ) (cos cos )n nkk r
bσ θ θϕ= + − , (11) 10
{ }' '
m m s 1 m 1 m( tan ) 1 exp( ( ) (1 )(sin sin ) / )c r s kτ σ ϕ θ θ θ θ = + × − − − − − − . (12)
When the wheel is in a quasi-static state, the effect of the distributed stress (normal stress σ and
tangential stress τ) can be simplified to the normal force FN, drawbar pull FDP, and driving resistance torque
MR by integrating along with the wheel–terrain interaction area, which are balanced with the wheel load W,
resistance force fDP, and driving torque T, respectively. For the wheels of Zhurong moving almost on flat 15
terrain with a maximum speed of 200 m/h, the quasi-static condition is valid because the dynamic effects
are negligible at low speeds. Therefore, the force/torque balance equations for Zhurong’s lugged wheels
can be expressed as follows:
( ) ( ) ( ) ( ){ }m 1
2 mN 2 s 2 1 s 1cos sin d cos sin d =F b r r r r W
θ θ
θ θσ θ θ τ θ θ θ σ θ θ τ θ θ θ= + + + ∫ ∫ , (13a)
( ) ( ) ( ) ( ){ }m 1
2 mDP s 2 2 s 1 1 DPcos sin d cos sin dF b r r r r f
θ θ
θ θτ θ θ σ θ θ θ τ θ θ σ θ θ θ= − + − = ∫ ∫ , (13b) 20
21
( ) ( )m 1
2 m
2
R s 2 1d dM r b Tθ θ
θ θτ θ θ τ θ θ = + = ∫ ∫ (13c)
According to Shibly et al.39, the wheel-terrain interaction stress can be linearized as follows:
( ) ( ) ( ) ( )( ) ( ) ( ) ( )
1 m 1 1 m m 1
2 m 2 m 2 2 m
= /
= /
σ θ σ θ θ θ θ θ θ θσ θ σ θ θ θ θ θ θ θ
− − ≤ ≤ − − ≤ ≤
, (14)
( ) ( ) ( ) ( )( ) ( ) ( ) ( )
1 m 1 1 m m 1
2 m 2 m 2 2 m
= /
= /
τ θ τ θ θ θ θ θ θ θτ θ τ θ θ θ θ θ θ θ
− − ≤ ≤ − − ≤ ≤
. (15)
Because the entrance angle is usually not large, and when θ approaches θ1, the corresponding normal 5
stress and tangential stress all approach zero, Ding40 proposed that the product of cosθ and stress can also
be linearized. Let m m m= cosσ σ θ′ and m m m= cosτ τ θ′ . Then, it can be deduced that
( ) ( ) ( )( ) ( ) ( )
1 m 1 1 m
2 m 2 m 2
cos = /
cos = /
σ θ θ σ θ θ θ θσ θ θ σ θ θ θ θ
′ − − ′ − −
, (16)
( ) ( ) ( )( ) ( ) ( )
1 m 1 1 m
2 m 2 m 2
cos = /
cos = /
τ θ θ τ θ θ θ θτ θ θ τ θ θ θ θ
′ − − ′ − −
. (17)
Bringing equations (16)~(17) into the wheel–terrain interaction model of equation (13a), and ignoring 10
the vertical component of the shearing stress, we can obtain simplified expressions40 of the normal force FN
and the driving torque MR as
( )N 1 2 m mcos / 2F rb θ θ σ θ≈ − , (18)
( )2
R s 1 2 m= / 2M r b θ θ τ− , (19)
while equation (18) and equation (19) can be rearranged as 15
Nm
1 2 m
2
( )cos
F
rbσ
θ θ θ=
−, (20)
22
Rm 2
s 1 2
2
( )
M
r bτ
θ θ=
−. (21)
Equation (20) is combined with equation (11) to obtain the relationship of bearing characterisitic
parameters (cohesive modulus ck , frictional modulus kϕ , and sinkage exponent n of the soil). When the
normal force FN and the wheel sinkage z are estimated while the wheel parameters are known, let
Ks=kc/b+kφ, then the bearing characteristics curves of the soil for the sinkage exponent n and the equivalent 5
stiffness modulus Ks under different wheel sinkages can be plotted as in Fig. 4a.
Regarding the analysis of the soil shearing parameters, we find equations (12) and (21) are two
expressions of the maximum tangential stress τm, and the maximum shearing stress τm can be directly
calculated according to equation (21), with the measured driving resistance torque MR. Then, all quantities
in equation (12) can be measured except for the three unknown shearing parameters (the cohesion c, the 10
internal friction angle φ, and shearing deformable modulus k). The shearing deformable modulus k is
determined by the slope of the shearing curve at the origin point and the maximum tangential stress τm. Its
value is 1/3 of the corresponding shearing deformation when the shearing stress τ is equal to 95% of the
maximum shearing stress τm. The denser the soil, the smaller the value of k. We concluded in the main text
that Martian soil at the Tianwen-1 landing site has a larger bearing strength than the lunar regolith. The 15
typical value of the shearing deformable modulus k of the lunar regolith is 17.8 mm, therefore it is assumed
here that the soil shearing deformable modulus k at the Zhurong landing site is 5 mm. Since the soil
parameters at the Zhurong landing site have not been rigorously identified, a value of k of 5 mm is not
accurate, but it can be used for inference. Therefore, the shearing characteristics curves of the soil for the
cohesion c and the internal friction angle φ under different driving torques can be plotted. 20
In practice, the values of the wheel radius r, wheel width b, wheel lug height h, and equivalent radius
rs are constant and determined by Zhurong’s wheel configuration, as shown in Extended Data Table 1. The
normal force FN is estimated from the vertical load W, which can be computed from a quasi-static force
23
analysis of the rover with knowledge of the rover configuration and mass distribution. The two key wheel
motion state indicators, the slip ratio s and sinkage z, can be computed using vision-based techniques or
kinematic analysis of the rover suspension. Based on the given parameters, for wheel sinkages of 2, 5, 10,
and 15 mm, the curves of the equivalent stiffness modulus (denoted by Ks) and the sinkage exponent n were
plotted (Fig. 4a). The motor current of the driving wheel varied from 0.17 A to 0.28 A with an average of 5
0.23 A when the wheel was rotating. Using the relationship between the motor current and the driving
torque for the driving wheel, as shown in Extended Data Fig. 7b and the reduction ratio of the reduction
drive between the wheel driving motor and the driving wheel (3.92×160 with an efficiency of 60%), the
driving torque was calculated to be 3.7~7.0 N∙m, and the average driving torque for each continuous steady
traverse could be calculated for analysis. The maximum and minimum average driving torques are 3.7 N∙m 10
and 7.0 N∙m, respectively. The average driving torque is mostly around 5.5 N∙m. Therefore, for driving
torques of 4, 5, 6, 7, and 8 N∙m, the curves of the cohesion c and the internal friction angle φ were plotted
(Fig. 4b).
References
32. Zheng, J. et al. Design and terramechanics analysis of a Mars rover utilizing active suspension. Mech. 15
Mach. Theory 128, 125-149 (2018).
33. Gao, H., Zheng, J., Liu, Z., Yu, H., Ding, L., Li, N., Deng, Z. Performance analysis on wheels lifting-
off-ground for Mars rover with active suspension, Robot 39, 139-150 (2017).
34. Ding, L., Gao, H., Deng, Z., Yoshida, K., Nagatani, K. paper presented in the 2009 IEEE/RSJ
International Conference on Intelligent Robots and Systems, St. Louis, MO, USA, 10 to 15 October 20
2009.
35. Li, N. thesis, Harbin Institute of Technology (2019).
24
36. Ding, L., Gao, H., Liu, Z., Deng, Z. & Liu, G. Identifying mechanical property parameters of planetary
soil using in-situ data obtained from exploration rovers. Planet Space Sci. 119, 121-136 (2015).
37. Bekker, M. G. Introduction to Terrain-Vehicle Systems, Part I: The Terrain. Part II: The Vehicle
(Michigan Univ Ann Arbor, Ann Arbor, MI, 1969).
38. Terzaghi, K., Peck, R. B., Mesri, G. Soil Mechanics in Engineering Practice (John Wiley & Sons, New 5
Jersey, 1996).
39. Shibly, H., Iagnemma, K., Dubowsky, S. An equivalent soil mechanics formulation for rigid 25 wheels
in deformable terrain, with application to planetary exploration rovers, J. Terramech. 42, 1-13 (2005).
40. Ding, L. thesis, Harbin Institute of Technology (2009).
Data availability: Datasets generated or analysed during this study are available from the corresponding 10
authors upon reasonable request.
Acknowledgments: This work was supported in part by the Development Program of China under Grant
2019YFB1309500, in part by the National Natural Science Foundation of China under Grant 51822502 and
Grant 91948202, in part by the ‘111 Project’ under Grant B07018, in part by the Heilongjiang Postdoctoral
Fund under Grant LBH-Z20136, in part by the Self-Planned Task of State Key Laboratory of Robotics and 15
System (HIT) under Grant SKLRS 202101A03, and in part by the Pre-research project on Civil Aerospace
Technologies by CNSA under Grant D020102.
Author contributions: L. Ding coordinated; and wrote the manuscript with R. Zhou. T. Yu led the
Zhurong rover operation and data acquisition. H. Gao and H. Yang coordinated coauthor contributions.
Data and image analysis: L. Ding, R. Zhou, H. Gao, H. Yang, Y. Yuan, Z. Wang, Z. Deng, L. Huang, N. Li, 20
Z. Li, F. Niu, H. Qi, S. Li, W. Feng, C. Yang, H. Xing, G. Wang, L. Niu, P. Xu. Rover operations and data
acquisitions: T. Yu, J. Li, C. Liu, J. Wang, X. Cui, X. He, H. Zhang, R. Zhao, Z. Zhang, Z. Cheng, F. Wu,
Q. Xu, H. Lu, L. Li, X. Wang, Z. Huang, J. Zhang. Rover parameters provision and analysis: Y. Jia, B.
25
Yuan, B. Chen, Z. Dang. Geological analysis and interpretation: Y. Zhao, X. Wang, G. Bao, W. Wan, M.
Zhang, K. Di. Results and writing refinement: G. Liu and L. Richter. All authors reviewed and revised the
manuscript.
Competing interests: The authors declare that they have no competing interests.
Additional Information: Supplementary Information is available for this paper. Correspondence and 5
requests for materials should be addressed to L. Ding (email: [email protected]), or H. Gao (email:
[email protected]) or H. Yang (email: [email protected]).
26
Extended Data Fig. 1. Zhurong rover on Tianwen-1 mission. a, The first group photo of Zhurong rover
and Tianwen-1 lander on the Martian surface. The image was taken by a wireless camera deployed by
Zhurong on sol 18 (June 1, 2021). b, Six scientific payloads and the lugged wheel on the Zhurong rover.
The six scientific payloads2 include NaTeCam, Multispectral Camera (MSCam)3, Mars Rover Penetrating 5
Radar (RoPeR)4, Mars surface Composition Detector (MarSCoDe)7, Mars Rover Magnetometer
(RoMAG)6, and Mars Climate Station (MCS)5. The lugged wheel is evenly arranged with 20 lugs on its
outer surface and spokes inside. c, Creeping and crabbing mode of Zhurong rover relied on active
suspension joints. The creeping mode of Zhurong to get out of sunk and climb steep slopes is realized by
decreasing or increasing the angle of the main rocker arm, coordinated with the movement of other wheels. 10
The crabbing mode of Zhurong for lateral movement is achieved by turning the steering wheels at 90° and
driving laterally.
27
Extended Data Fig. 2. HiRISE image of Tianwen-1 probe. The HiRISE image is ESP_06937_2055 at
29.2 cm/pixel (with 1×1 binning) acquired by NASA’s Mars Reconnaissance Orbiter on June 2, 2021. a, A
regional view of the location of the Tianwen-1 lander, Zhurong rover, parachute and backshell, and
heatshield. The backshell and parachute are located at 109.923° N, 25.059° E, ~350 m (Northing = 330.6 5
m, Easting = 114.7 m) in the southwest of the lander. The heatshield is located at 109.901°N, 25.048°E,
~1617 m (Northing=1307.4 m, Easting=944.5 m) southwest from the lander. In the close up color image,
the large bright spot is the lander of Tianwen-1, while the small bright spot is the Zhurong rover. There are
symmetrical bright radial streaks of jets in the north-south direction centered on the lander, which might be
due to the engine plume during landingblowing away the original surface materials. The backshell is of the 10
round bright spot and the parachute is of the long bright spot. The close up of the heatshield is shown in the
panchromatic image. b, Distribution and characteristics of dunes around the Tianwen-1 landing site on
Mars. The three enlarged views show three types of sand dunes in shapes: the seagull shape, the crescent
shape and the line shape. Most dunes in this area are in a crescent shape and their windward slopes are
(yellow arrows) oriented to the NE-SW direction. 15
28
Extended Data Fig. 3. Rover locomotion simulation. Simulation of four traverses after the rover driving
off the ramp is visualized. The red line represents the planned path and the blue line represents the path
obtained in simulation. The movement sequence of the rover is carried out in the order of the number
marked. The terrain map is built from NaTeCam images. 5
29
Extended Data Fig. 4. Wheel track analysis. a, Stagged pattern of the wheel track left by a wheel moving
with longitudinal slip. b, Wheel track extraction and wheel slip ratio analysis over a traverse to waypoint A.
Eight continuous track units are extracted on the orthograph for local wheel slip ratio calculation. The
image was taken by the NaTeCam on sol 11. c, Near-Side view of the wheel-terrain interaction. Only wheel 5
lugs are submerged into the soil and the wheel rims of these three right wheels are millimetres above the
surface. Some soil adhered to the wheel surface or on the groove bordering the lugs. This image was taken
by the wireless camera on sol 12 when the rover was retreating. d, e, f, show wheel tracks and the
associated wheel sinkage. Images were taken by the rear HazCam. d, A part of legible wheel track with
well-trimmed contours. The wheel sinkage is estimated to be about 10 mm. e, Most common form of wheel 10
tracks without well-trimmed contours. Its wheel sinkage is estimated to be ~5 mm. f, shows both wheel
tracks interrupted by gravels and wheel tracks with rim sinking below the surface. The wheel sinkage of the
wheel tracks interrupted by gravels is estimated to be ~2 mm. The wheel sinkage of wheel tracks formed by
the wheel rim sinking below the surface is estimated to be ~15 mm.
15
30
Extended Data Fig. 5. Craters around the Tianwen-1 landing site. a, The distribution of the mapped
craters (diameters > 1 m) in the circular region surrounding the Tianwen-1 landing site, overlaid on the
HiRISE image (ESP_06937_2055). C1, C2, C3 are three large craters (diameter > 200 m) in this area. The
circular region is centred on 109.925° E, 25.048° N with a diameter of 3 km, and its centre is 1 km south 5
away from Tianwen-1 landing site. b, A log-log plot of the incremental size-frequency distribution of
craters. The diameter interval is 2 D m and the crater diameter refers to the middle value of each bin. c, A
small crater surrounded by rocks in dark color with sand deposited at its bottom. The image was taken by
NaTeCam on sol 34 (June 17, 2021). d, A crater suffered severe erosion, showing severely damaged rims
and lose clear impact structure but present as the depression. The image was taken by NaTeCam on sol 57 10
(July 11, 2021).
31
Extended Data Fig. 6. Rocks around the Tianwen-1 landing site. a,b,c, are rocks riddled with small
dense pits on the surface. The rocks appear relatively light in tone through the exposed surface with less
dust and soil covering. d,e, are rocks showing layered structures, and about three layers (divided by red
dash lines) can be seen on the sides. The rock flakes of the top layer and the middle layer seem different in 5
direction. f, shows rocks with one windward face full of grooves, likely worn and shaped by wind-blown
particles.
32
Extended Data Fig. 7. Wheel-soil interaction model and the driving motor characteristics curve. a,
Force diagram for the wheel-soil interaction of a lugged wheel. b, Characteristics curve of the driving
torque and the motor current for the driving motor on the Zhurong rover.
5
33
Extended Data Table 1. The parameters of the Zhurong rover.
Group Parameter name Value
Rover
Mass m (kg) 240
Size l×w×h (m) 1230×830×540
Number of wheel 6
Maximum driving speed vmax (m/h) 200
Maximum climbable slope on rigid ground (°) ≥30
Maximum climbable slope on soft ground (°) ≥20
Climbable obstacle (mm) ≥300
Lifting range of the rover body (mm) 0~500
Wheel
Maximum wheel radius rmax (m) 0.145
Minimum wheel radius rmin (m) 0.135
Wheel width b (m) 0.20
Height of the wheel lug h (m) 0.005
Number of wheel lugs 20
Shearing radius of a wheel rs (m) 0.146
Normal force on each wheel FN (N) (under quasi-static state)
148.8
Wheel slip ratio s -0.1~0.2
Wheel sinkage z (m) 0.005~0.01
Six wheels on the Zhurong rover can both driving and steering. The wheel’s largest radius (145 mm) is in the middle cross-
section and the smallest radius (135 mm) is on the two end faces at both sides. The 20 wheel lugs are evenly attach to the outer
edge of the wheel.
5
34
Extended Data Table 2. Daily scientific exploration events. Zhurong carried out routine navigation and
detection using NaTeCam, Mars rover penetrating radar (RoPeR), and Mars climate station (MCS) on most
sols. Scientific payloads2 like multispectral camera (MSCam), Mars surface composition detector
(MarsSCoDe), and Mars rover magnetometer (RoMAG) are used for a specific target detection, like rocks
and sand dunes. The sol represents a Martian day, corresponding to 24.65 hours at the early stage of the 5
mission after landing.
Sol Date Accumulated mileage
(m) Key events
1 May 15 0 Landing successfully at Utopia Planitia
5-8 May 19-22 5.48 Panoramic imaging and drive off the ramp
9-11 May 23-25 12.24
Panoramic imaging and discovery of parachute and backshell Wheel track imaging by the rear hazard avoidance camera Coarse landing site localization based on vision Wind measurements Moving to waypoint A with routine navigation and detection
12-16 May 26-30 18.73
Moving to waypoint B with routine navigation and detection Imaging the Chinese flag on the lander by Zhurong’s TeNaCam Two descent images are transmitted back to Earth Dropping the wireless camera and take test shots
18 June 1 27.16
Group imaging of the lander and the rover Fine landing site localization based on descent images High resolution imaging of the orbiter observed the lander, rover, parachute, backshell and heatshield Routine navigation and detection
22 June 5 37.19 Multi-spectral imaging on a rock full of holes Surface composition detection on a rock and the soil Routine navigation and detection
23-44 June 6-28 238.42 Switching the rover locomotion mode to high efficiency mode Moving toward north to find the parachute and backshell with routine naviation and detection
45-46 June 29-30 241.02 Moving approach to a sand dune and taking near-field images with routine navigation and detection Sand dune in-situ detection using the multi-spectral camera and the surface composition detector
48 July 2 251.07 Moving toward north to find the parachute and backshell with routine navigation and detection
49 July 3 271.09 Observation of a special rock
51-57 July 5-11 410.02 Moving toward north to find the parachute and backshell with routine navigation and detection
59 July 13 430.72 Group imaging the parachute and backshell
60 July 14 450.86 Moving to a nearby sand dune with routine navigation and detection
35
Extended Data Table 3. Driving motor current and wheel driving torque of traverses on sol 23-34.
Earth Data Sol Drive Distance (m) Driving Motor Current (A) Wheel Driving Torque (N∙m) Average Wheel Slip Ratio Average Maximum Average Maximum
June 6, 2021 23 5.03 0.28 0.35 7.0 9.1 0.02 June 8, 2021 25 5.30 0.17 0.22 3.7 5.0 0.07 June 10, 2021 27 8.12 0.25 0.33 5.9 8.5 0.06 June 11, 2021 28 9.13 0.24 0.39 6.0 10.3 0.07 June 12, 2021 29 9.64 0.25 0.41 6.1 10.9 0.06 June 13, 2021 30 9.82 0.20 0.28 4.6 6.8 0.05 June 15, 2021 32 8.83 0.22 0.33 5.0 8.5 0.09
The wheel driving torque is transformed from the driving motor current, considering the reduction rate and
efficiency.
Supplementary Files
This is a list of supplementary �les associated with this preprint. Click to download.
SupplementaryInformation.docx